Your Backtest Filled at the Price It Wanted: Slippage, Gaps and the Execution Gap Nobody Prices In
9 September 2026, 08:00
0
26
In the Strategy Tester, your order is filled at the price the model says. In a live account it is filled at the price someone is willing to give you, at that moment, in that liquidity. The distance between those two is a cost, and most published backtests set it to zero.
Where the gap comes from
Three sources dominate. Slippage: the market moves between the decision and the fill, and stop orders in fast markets are especially exposed. Weekend and news gaps: price reopens past your stop level, so the stop executes worse than it was set. Requotes and rejections on some execution models, which turn a planned trade into no trade at all.
It hits stops harder than entries
Entry slippage is roughly symmetric over many trades — sometimes better, sometimes worse. Stop-loss slippage is not: gaps tend to blow through the stop in the losing direction, because that is the direction the market was already moving. A backtest that assumes exact stop fills systematically understates the size of the worst losses.
Which strategies are exposed
The cost scales with how tight the target is relative to the noise. A scalper aiming for a few points can have its entire edge consumed by half a point of slippage per trade. A daily-timeframe system aiming for a multi-week move can absorb the same slippage without noticing. Frequency multiplies it: 2,000 trades a year at one extra point each is a real annual charge.
How to test for it
Re-run the backtest with a wider spread than your broker's average and, if your platform allows it, with an added slippage assumption. Then compare the result to the original. If a strategy only works with perfect fills, it does not work.
How we handle it
We run our published tests on real tick data with real variable spread rather than an idealised fixed-spread model, because a fixed-spread test quietly removes the exact moments — news, rollover, thin liquidity — when execution is worst. That does not simulate your broker's slippage, so we also tell buyers to confirm behaviour on a demo account with their own broker first.
Our three best-selling EAs: GOLD NEURON (AI) https://www.mql5.com/en/market/product/187329 · ATLAS PORTFOLIO https://www.mql5.com/en/market/product/182751
Every figure we publish comes from real tick data with real spread, not from an idealised fill model. The full list: fxea365.com/ea/ranking


